Q: OSX Yosemite Wifi issues

MacBook pro retina mid 2014. Same problem with wifi as others. Tried to turn of bluetooth and it worked. From 1-2 mb to 47 mbit/s, simple as that. I have to add that I also changed two other settings first mentioned here but they did not work.

Weird, is it not? I feel I am experiencing the opposite. I bought the Apple Battery Charger kit because I got so tired of changing alkaline batteries so frequently. Best 30 bucks I've ever spent! They last way longer, hold a better charge and the kit comes with 6 ea. AA batteries. So, my first set of batteries got below 20% and I started seeing a drop in the Wi-Fi TX Rate. Usually running around 405 to 450. When the battery % level dropped to 20% and below my TX Rate dropped to between 200 and 300 never going above. I'm only using a Magic Trackpad on my Mini and my Full Apple Keyboard is wired. That's why I asked the same question as you. Fact or coincidence. Have we become hypersensitive? So, it would be interesting to know, but it may be too much info to know the ins and outs of the Broadcom card and interaction of Bluetooth and Wi-Fi, thereof.

This afternoon I changed the batteries on both. trackpad was at 5% and Keyboard was at 3%. Right after changing the batteries went to download a file. I watched it slowly decrease in speed from 3.1GB down to 250kbs. Knew it wouldn't finish so stopped it and rebooted. Once rebooted web pages taking longer to load, wifi froze had to reboot. Decided to take the new batteries out and put the old ones back in. Have not had an issue.

This makes pretty much sense to me. With low battery, I suspect the BT radio transmission/ antenna power to be lower. Thence possibly less interference on the 2.4 GHz band. Try using rechargeable NiMh cels. Their voltage is slightly lower (1.3V) than alkaline cells (1.5V).

On the other hand, JnDupuis in the post above has the opposite story. Which puzzles me, in fact. The only explanation for that would be that with low battery power, the BT/WiFi chip has to try to connect to BT so very often that WiFi discovery gets lagged. Unlikely theory, but well...

This BT/WiFi interference issue is a strange creature anyhow. It seems to vary a lot between different Mac models. Could be a different chip, could be a different (triple) antenna design. Remember: radio transmission is always a bit voodoo, even today.

PFJ30 wrote:

 

And what, may I ask, was so awful about my etre check result, that no-one has dared comment on it?!

Nothing afwul Paul, just some awkwardness that needed more research. Didn't have much time for research last night.

Johns Etrecheck needed no research whatever.

A few things:

- about your Time Machine disk being too small, you probably know.

- Since Mavericks it is highly unlikely to find page-outs (ram content being bounced to HDD, aka Virtual Memory, aka ram swapping). Your Etrecheck does show some (313 MB). Wondering why... In particular with the new memory management in Yosemite, swapping should completely be a thing of the past.

- 10 GB of page-ins is enormous. That is an improbably high amount. Firt thing that needs investigating

It is voodoo Hex, RF is either following the path of least resistance or attracted to as much noise as possible. Hey! how about an antennae port and plug-in antennae to get the Wi-Fi antennae out of the Mac case? Sounds retro, but we're finding today some of the old was more reliable than we thought. I mean, come on! An RF antennae broadcasting directly around the CPU and Board? Really? 

PFJ30 wrote:

 

And what, may I ask, was so awful about my etre check result, that no-one has dared comment on it?!

Nothing afwul Paul, just some awkwardness that needed more research. Didn't have much time for research last night. Sorry!

Johns Etrecheck needed no research whatever.

A few things:

- in general looked like a rather neat report, but with a few puzzling anomalies which need further evaluaton.

-" [failed]    com.brother.LOGINserver.plist": this may be about some antiquated software that came with your printer; it being only a preference list and it not loading anyhow, I suggest you thrash it and restart. Preference lists will auto-rebuild themselves anyhow.

- "InstUtilLaunch    UNKNOWN  (missing value)" needs investigation. Seems corrupt. Probably a remnant of an earlier migration.

- about your Time Machine disk being too small, you probably know.

- Since Mavericks it is highly unlikely to find page-outs (ram content being bounced to HDD, aka Virtual Memory, aka ram swapping). Your Etrecheck does show some (313 MB). Wondering why... In particular with the new memory management in Yosemite, swapping should completely be a thing of the past.

- 10.39 GB of page-ins is enormous as well as unlikely. That is an improbably high amount. First thing that needs investigating & fixing ASAP! Any page-in larger than the physical amount of ram in your Mac should theoretically be impossible. Paul, please check Activity Monitor to see if it is also stating these excessive page-in numbers!

- Last 2 lines of the report: at least those 2 diagnostics reports need to be looked into.

So now you know what I've been trying to figure out. Still at it as I type.

Well, Etrecheck points you to possible issues, the user or advisor is left with defining the right questions and investigating those... Fine, I'm investigating

jndupuis1 wrote:

 

It is voodoo Hex, RF is either following the path of least resistance or attracted to as much noise as possible. Hey! how about an antennae port and plug-in antennae to get the Wi-Fi antennae out of the Mac case? Sounds retro, but we're finding today some of the old was more reliable than we thought. I mean, come on! Really? 

Some of the old was indeed more reliable than we thought. The old knew their physics, the young always seem to think adding another microprocessor will do the trick.

Heck, I've even seen some coffee machines with a microprocessor on the mains breaker board- just meant to switch mains on or off

An RF antennae broadcasting directly around the CPU and Board?

Not to mention broadcasting with antennae in an aluminum (for the Brits here: aluminium) Faraday cage... Physics, Mr. ives, physics.

Exactly!!! Let's add to it by broadcasting higher frequency as well......Ladies and Gentlemen a warm hand for - 5 GHz. Let's give it up for Wireless AC as well. We'll render our CPU's and Motherboards useless in an Aluminum Case. Seriously, though, put a Transistor Radio in an MRI machine...will there be RF interference? How about a Transistor Radio on top of the old PAL or NTSC TV Set...will there be RF interference? So Broadcom nailed it by putting the Wi-Fi and Bluetooth chip next to each other on their board. Then got Apple to buy their Lemon at a wink, wink "you're gonna love this deal price." Couple of fruits, eh?

2.4 GHz is going to become extinct. We'll eventually be using 5 GHz old technology in a Wireless AC world. The days of internal antennae on Laptops, Portable Devices and Desktops are going to have to improve...Or insulating high RF broadcast from internal CPU components will have to improve.

Okay, I know, getting off topic.

The post was even bigger; ran into these edit timers again:

jndupuis1 wrote:

 

It is voodoo Hex, RF is either following the path of least resistance or attracted to as much noise as possible. Hey! how about an antennae port and plug-in antennae to get the Wi-Fi antennae out of the Mac case? Sounds retro, but we're finding today some of the old was more reliable than we thought. I mean, come on! Really? 

Some of the old was indeed more reliable than we thought. The old knew their physics, the young always seem to think adding another microprocessor will do the trick.

Heck, I've even seen some coffee machines with a microprocessor on the mains breaker board- just meant to switch mains on or off

An RF antennae broadcasting directly around the CPU and Board?

Not to mention broadcasting with antennae in an aluminum (for the Brits here: aluminium) Faraday cage... Physics, Mr. Ives, physics.

And some other historic issue for the tech- savvy: "tin whiskers", the cause of much grief with lead-free solder... Adding some lead to solder was introduced  in the 50ies specifically to avoid those tin whiskers. Nowadays the use of lead is forbidden, but this is causing a far greater failure rate of electronic appliances. Which of either old or new options is best for the environment? I'd opt for the elder option...

Sorry for the offtopic.

Tried restoring mDNSdiscovery from 10.10.2. Same result, no discernible difference.

I didn't combine it with any other changes (such as changing the kext, destroying plists, rebuilding networking locations etc.)

This shouldn't be this hard.

But I still like the theory that my Mac under 10.10.3 (and probably somewhat under 10.10.2) is noisy, and public networks (big hotel, large corporate guest WiFi) simply 'shut it out'. Of note, I am fairly certain the hotel in question (JW Marriott) and my client corporate guest network both use the same Cisco network hardware/software/access control.

Other networks (my home, small coffee shops) seem to work fine.

Matthew

Right. Back full on-topic: would any of the Yosemite plaintiffs here please check their Activity Monitor and look for any suspiciously high "Page-in" numbers showing at the bottom of that window?

Meaning numbers going above your Macs physical RAM capacity?

Would you next download and run Etrecheck (http://www.etresoft.com/etrecheck) and at the bottom of the report see if the Virtual Memory "Page-in" numbers correlate with those of Activity Monitor? If they do please report here.

I'm trying to find a Yosemite bug , and may be close to finding it. Thank you!

mmontano73 wrote:

 

Tried restoring mDNSdiscovery from 10.10.2. Same result, no discernible difference.

This should read substituting with mDNSresponder from Mavericks 10.9.5! Please retry!

 

I didn't combine it with any other changes (such as changing the kext, destroying plists, rebuilding networking locations etc.)

You should, methodically and step-by-step.

 

This shouldn't be this hard.

No, but sometimes it seems to be. More than 2,755 Replies later in this thread, still a mystery.

 

But I still like the theory that my Mac under 10.10.3 (and probably somewhat under 10.10.2) is noisy, and public networks (big hotel, large corporate guest WiFi) simply 'shut it out'. Of note, I am fairly certain the hotel in question (JW Marriott) and my client corporate guest network both use the same Cisco network hardware/software/access control.

Agreed, and an interesting testimony to boot;; try to switch off Airdrop and and Handoff while at a hotel and see how you fare.

 

Other networks (my home, small coffee shops) seem to work fine.

Your issue may lie within the Yosemite + WPA Enterprise sub-issue, maybe?

 

 

Good luck!

I see there are many responses including possible solutions to this issue, but ultimately this needs to be fixed for the average user. Is there an update to fix this yet?

I don't have the time nor technical expertise to fix this myself and a wired connection is not an option for my MacBook Air. I just need to be able to use my computer.

Thanks!

(posted from a Windows PC)

https://twitter.com/dairymews/status/571665378155347969

A personal message to Tim Cook.

I knew I Iwasnt abandoned  

POlitical canvassing AM, Photography Exhib PM, your flagged areas and my new bootable 16g flash drive await this evening